Infrared spectra information and their correlation with QSAR descriptors

In principle, the InfraRed (IR) spectra have very appealing properties for QSAR research: they are generated in the range of low energy molecular interactions that play a fundamental role in life (e.g., for molecular recognition), and they are extremely specific fingerprints of the molecules. We compared the information carried by the fingerprint region of the IR spectra (1500-600 cm(-1)) with that of a range of descriptors presently in use in the QSAR field: (a) classical physical chemical and quantum mechanical properties (logP, MR, HOMO, LUMO); (b) molecular connectivities; (c) 2-D molecular distances; and (d) a novel infrared range vibration based theoretical descriptor (EVA). Much redundancy and overlapping was found among descriptors such as connectivities, 2-D distances, and theoretical spectral EVA descriptors. On the contrary, the complex information carried by IR spectra (fingerprint region) was markedly different from that codified in various molecular descriptors presently in use in QSAR practice, thus pointing to the importance of further studying the potential relevance of IR information for QSAR analysis.